PARP1-driven repair of topoisomerase IIIα DNA-protein crosslinks by FEN1

SUMMARY

Persistent DNA-protein crosslinks formed by human topoisomerase IIIα (TOP3A-DPCs) interfere with DNA metabolism and lead to genome damage and cell death. Recently, we demonstrated that such abortive TOP3A-DPCs are ubiquitylated and proteolyzed by Spartan (SPRTN). Here, we identify transient poly(ADP-ribosylation) (PARylation) in addition to ubiquitylation as a signaling mechanism for TOP3A-DPC repair and provide evidence that poly(ADP-ribose) polymerase 1 (PARP1) drives the repair of TOP3A-DPCs by recruiting flap endonuclease 1 (FEN1) to the TOP3A-DPCs. We find that blocking PARylation attenuates the interaction of FEN1 and TOP3A and that TOP3A-DPCs accumulate in cells with compromised PARP1 activity and in FEN1-deficient cells. We also show that PARP1 suppresses TOP3A-DPC ubiquitylation and that inhibiting the ubiquitin-activating enzyme E1 (UBE1) increases TOP3A-DPCs, consistent with ubiquitylation serving as a signaling mechanism for TOP3A-DPC repair mediated by SPRTN and TDP2. We propose that two concerted pathways repair TOP3A-DPCs: PARylation-driven FEN1 excision and ubiquitylation-driven SPRTN-TDP2 excision.

Graphical abstract

In brief

Saha et al. report a regulatory mechanism of TOP3A-DPC repair coupling FEN1 and PARP1 to maintain genome integrity. The PARylation of TOP3A-DPCs by PARP1 primes their repair by FEN1. Transient PARylation of TOP3A-DPCs suppresses their ubiquitylation, and vice versa, which highlights the crosstalk between two post-translational modifications.

INTRODUCTION

Topoisomerases comprise a family of six enzymes that eliminate topological stress and entanglements of nucleic acids arising from fundamental nuclear transactions.¹ They all act by a common mechanism of cleavage and rejoining of DNA/RNA strand(s) using a catalytic tyrosine to break and reseal the nucleic acid phosphodiester backbone. Hence, topoisomerase cleavage complexes (TOPccs) are the key catalytic intermediates for all topoisomerase reactions,² which normally need to be transient and readily self-reversible in nature. Yet, TOPccs can be stabilized or trapped on DNA (persistent TOPccs) due to abortive catalytic cycles that are referred to as topoisomerase DNA-protein crosslinks (TOP-DPCs).³ TOP-DPCs can be formed endogenously by DNA alterations such as abasic sites, mismatches, alkylated bases, carcinogenic adducts, and DNA breaks, leading to a failure in completion of the topoisomerase enzymatic reactions.^1,4 TOP-DPCs are also targeted by widely used anticancer drugs trapping TOP1ccs and TOP2ccs^5,6 and antibacterial drugs targeting gyrase and Topo IV.^7–9

Topoisomerase IIIα (TOP3A) is a type IA topoisomerase. It cleaves single-stranded segments within duplex DNA to pass another DNA strand through the break, reseals the ends of the broken DNA, and dissociates from DNA in its normal catalytic cycles.¹⁰ Thus, TOP3A forms TOP3A-linked single-strand breaks (SSBs; TOP3Accs) with a TOP3A molecule covalently bound to the 5′ end of the break.^3,11 TOP3Accs can be examined by mutating the enzyme catalytic site (R364W), which generates persistent TOP3A-DPCS,¹¹ and cells must excise the TOP3A-DPCs, which otherwise may cause DNA damage, genomic instability, and cell death.¹¹

DNA SSBs and/or gaps derived from unligated Okazaki fragments form in nascent DNA strands behind replication forks,¹² and TOP3A has been proposed to act on the precatenanes or DNA structures containing the Okazaki fragments in the lagging strand.^11,13 The short 5′ flap generated at the junction of single-stranded DNA (ssDNA) and double-stranded DNA when DNA polymerase displaces the strands during lagging strand synthesis is also effectively recognized and cleaved by the structure specific flap endonuclease 1 (FEN1), a member of the 5′ nuclease superfamily including yeast Rad2.¹⁴ This allows the processing of Okazaki fragments and the formation of long uninterrupted newly replicated DNA strands.¹⁵ Unligated Okazaki fragments also form ssDNA gaps that are sensed by poly(ADP-ribose) polymerase 1 (PARP1), a prominent DNA damage response (DDR) enzyme that binds to DNA lesions and catalyzes the attachment and polymerization of ADP-ribose chains to proteins (poly(ADP-ribosylation) [PARylation]) in the vicinity of DNA breaks.¹⁶ Recent studies showed that unligated Okazaki fragments are a major source of PARP1 activity in S-phase cells,^16,17 suggesting a potential link between FEN1, PAR, and TOP3A. Because we recently discovered the importance of Spartan (SPRTN), TDP2, and MRE11 in TOP3A-DPC repair in human cells,¹¹ we explored whether FEN1 is an additional cellular mechanisms to remove TOP3A-DPCs associated with replication and how it is coordinated with the SPRTN-TDP2 pathway.

Post-translational modifications (PTMs) such as ubiquitylation, SUMOylation (the conjugation of small ubiquitin-like modifiers [SUMOs] to lysine residue of target proteins), and PARylation are key regulators of repair pathway choice for TOP1- and TOP2-DPCs.^3,4,18 PARylation, one of the most common and widespread PTMs at DNA breaks/lesions, is catalyzed by PARPs that attach negatively charged poly(ADP)ribose (PAR) molecules and polymer chains to target proteins and itself.¹⁹ PARylation has been implicated in a wide range of biological processes including DNA repair, chromatin remodeling, DNA damage response (DDR), inflammation, transcription, apoptosis, and mitosis. PARylation has been extensively studied for its key roles in the DDR and DNA damage repair pathways, and PARP inhibitors (PARPis) have recently become widely used for the treatment of cancers.²⁰ PARP1 and PARylation provide a scaffold for the recruitment of various chromatin remodelers, DNA repair proteins, and nucleases to DNA damage sites, thereby facilitating DNA repair.^21–24 Yet, excessive accumulation of PAR (hyperPARylation) in chromatin induces genomic instability and/or cell death,²⁵ and PAR polymers need to be degraded or released from chromatin to maintain cellular homeostasis.²⁶ In cells, PAR polymers are primarily degraded by poly(ADP-ribose) glycohydrolase (PARG), which cleaves the unique 2′,1′-glycosidic ribose–ribose bonds of the PAR chains and releases free ADP ribose moieties. Thus, the synthesis and degradation of PAR chains are transient, with a very short half-life, and dynamically controlled in vivo.²⁷ In the context of TOP-DPCs, we recently found that PARylation of TOP1-DPCs blocks their degradation by the proteasome and regulates the ubiquitylation of TOP1.¹⁸

SUMOylation and ubiquitylation are also key regulators for the repair of a broad range of DPCs including TOP1- and TOP2-DPCs.^28,29 Yet, whether TOP3A-DPCs are subjected to PARylation has so far not been explored. In the present study, we found that PARP1 PARylates TOP3A-DPCs and investigated whether TOP3A-DPC PARylation serves as a signaling mechanism for TOP3A-DPC repair and how this modification is interconnected with the ubiquitylation-mediated proteolytic repair of TOP3A-DPCs by SPRTN.¹¹

RESULTS

FEN1-mediated repair of TOP3A-DPCs

Previous biochemical studies have shown that FEN1 is involved in the repair of TOP2-DPCs containing 5′-phosphotyrosyl termini.³⁰ Accordingly, FEN1-deficient vertebrate cells are hypersensitive to the TOP2 poison etoposide.³¹ Because TOP3A-DPCs also contain 5′-phosphotyrosyl termini, we reasoned that FEN1 might excise TOP3A-DPCs. To induce and detect cellular TOP3A-DPCs, we transfected human cells with our recently designed self-trapping TOP3A mutant (TOP3A-R364W)¹¹ and performed rapid approach to DNA adduct recovery (RADAR) assays to isolate DNA covalently bound protein adducts and detect TOP3A-DPCs (Figure 1A).¹¹

Figure 1. Accumulation of TOP3Accs in FEN1-deficient human cells.

(A) Top: outline of the experimental protocol for transfection followed by RADAR assay in MCF-7 cells with and without FEN1 inactivation and ectopic expression of TOP3A-R364W. Bottom (left): representative slot blot for TOP3Acc detection by RADAR assay from cells with the indicated genotypes. TOP3Accs were detected with anti-TOP3A antibody. Bottom (right): bar graph from quantitation of the 3 independent experiments as shown on the left. Error bars represent mean ± SD. p values were obtained from two-tailed unpaired t test with Welch’s correction. *p < 0.05.

(B) Left panel: representative slot blot detecting endogenous TOP3Accs in shControl and shFEN1 MCF-7 cells. TOP3Accs were detected by RADAR assay with anti-TOP3A antibody. Right: quantitation of TOP3Accs from three independent RADAR assays as shown on the left. Error bars represent mean ± SD. p values were obtained from two-tailed unpaired t test with Welch’s correction. **p < 0.01.

(C) Top: outline of the experimental protocol in HCT116 cells expressing TOP3A-R364W. Cells were transfected with either siControl or siFEN1 for 72 and 24 h with TOP3A-R364W for 48 h before performing RADAR assay to detect TOP3A-DPCs. Bottom (left): representative slot blot for TOP3Acc detection by RADAR assay from siControl and siFEN1 cells. TOP3Accs were detected with an anti-TOP3A antibody. Bottom (right): bar graph from quantitation of the 3 independent experiments as shown on the left. Error bars represent mean ± SD. p values were obtained from two-tailed unpaired t test with Welch’s correction. **p < 0.01.

(D) Top: HCT116 cells were transfected with the TOP3A-R364W expression plasmid for 48 h, followed by treatment with either FEN1i or DMSO (drug solvent) (10 μM, 1 h) before cell harvest. Cells were then subjected to the RADAR assay to detect TOP3A-DPCs by slot blot. DMSO was used in the control samples. Bottom (left): representative slot blot of RADAR assay from control and FEN1-inhibited cells. Bottom (right): bar graph from quantitation of the 3 independent experiments as shown on the left. Error bars represent mean ± SD. p values were obtained from two-tailed unpaired t test with Welch’s correction. **p < 0.01.

(E) Ectopic expression of FEN1 reduces TOP3Accs. Top left: outline of the experimental protocol. Bottom left: representative slot blot for MCF-7 cells first transfected with TOP3A-R364W and, 1 day later, transfected with FEN1-WT or the nuclease-dead mutant FEN1 (FEN1-D118A) plasmids. Cells were transfected in 6-well plates with 1.2 μg of each of the TOP3A-R364W and FEN1 plasmids (1:1 ratio). Representative RADAR assays with an anti-TOP3A antibody were used for TOP3A-DPC detection. EV, empty vector transfected. Right: bar graph from quantitation of 3 independent experiments as shown on the left. Error bars represent mean ± SD. p values were obtained from two-tailed unpaired t test with Welch’s correction. ***p < 0.001 (EV vs FEN1-WT) and ns: not significant (EV vs FEN1-D181A).

In (A)–(E), the indicated amounts of DNA were loaded, and loading was further tested by slot blot and probing with an anti-double-stranded DNA (dsDNA) antibody.

To evaluate the role of FEN1 in TOP3A-DPC repair, we depleted FEN1 by expressing small hairpin RNA (shRNA) against FEN1 in human breast cancer MCF-7 cells (Figure S1A) and measured TOP3A-DPCs by RADAR assay. We found that the level of TOP3A-DPCs was higher in FEN1 knockdown cells than in control shRNA cells (Figure 1A). Notably, FEN1-depleted (shFEN1) cells also displayed detectable endogenous TOP3A-DPCs (Figure 1B). To confirm the implication of FEN1 in TOP3A-DPC removal, we silenced FEN1 by small interfering RNA (siRNA) in human colon carcinoma HCT116 cells transfected with TOP3A-R364W (Figure S1B) and observed increased TOP3A-DPCs in FEN1-deficient cells in comparison with siControl cells (Figure 1C). Blocking FEN1 endonuclease activity with the small-molecule inhibitor FEN1-IN-4 (FEN1i)³² also increased TOP3A-DPCs in HCT116 cells expressing TOP3A-R364W (Figure 1D). Conversely, ectopic overexpression of wild-type (WT) FEN1 (Figure S1C) reduced TOP3A-DPC levels (Figure 1E), while the nuclease-dead FEN1 (FEN1-D118A)^33,34 (Figure S1C) was inefficient in reducing TOP3A-DPC levels (Figure 1E). These results suggested that FEN1 acts as a nuclease to excise cellular TOP3A-DPCs.

To examine whether the increased TOP3A-DPCs are associated with an increase in SSBs, we performed alkaline comet assays. Cells with ectopic TOP3A-R364W expression exhibited increased SSBs (tail moment) in comparison with mock-transfected cells with empty vector (EV) expression. Additionally, FEN1 depletion (shFEN1) in TOP3A-R364W-expressing cells caused a further increase in SSBs (Figures S1D and S1E). These data show that the elevated TOP3A-DPCs in the absence of FEN1 are associated with increased SSBs.

TOP3A-DPCs activate PARP1, and PARylation mediates the interaction of FEN1 with TOP3Accs

Next, we sought to identify the molecular mechanisms linking FEN1 with TOP3A-DPCs. First, we asked whether FEN1 interacts closely with TOP3A-DPCs and employed proximity ligation assays (PLAs) to check the FEN1 and TOP3A-DPC interaction using FEN1 and FLAG antibodies. PLAs using both FEN1 and FLAG antibodies revealed that cells transfected with FLAG-tagged TOP3A-R364W displayed PLA foci, while cells with no transfection did not, indicating that ectopically expressed FLAG-tagged TOP3A interacts with endogenous FEN1 (Figure 2A). Controls either without antibodies or with a single antibody did not form PLA foci (Figure S2A), validating the specificity of our PLA results.

Figure 2. PAR-mediated FEN1 recruitment and interaction of PAR with TOP3Accs in chromatin.

(A) PARP inhibition reduces FEN1 interaction with TOP3A. Top: protocol for the PLA experiments. Samples were prepared from U2OS cells non-transfected (NT) or transfected with FLAG-tagged TOP3A-WT or TOP3A-R364W for 48 h. Cells expressing FLAG-tagged TOP3A-R364W after 48 h transfection were treated with 1 μM talazoparib (PARPi) for 1 h before cell fixation and PLA. Bottom: representative images of a PLA for FLAG-tagged TOP3A and endogenous FEN1 for indicated conditions. Scale bar, 5 μm. Right: quantitation of the TOP3A-FLAG and FEN1 PLA foci shown in (A). Results are represented as SuperPlots³⁵ showing individual data points in different colors, the average from each experiment, and the mean of the averages ± SD. ****p < 0.0001, ordinary one-way analysis of variance (ANOVA) with Dunnett’s multiple comparison test.

(B) Representative FLAG immunoprecipitation (IP) assay showing the interaction of TOP3A with FEN1 and PARP1. Following transfection of FLAG-tagged empty vector (EV) or TOP3A-WT and TOP3A-R364W, U2OS cells were treated with either DMSO (drug solvent control), PARGi or PARPi alone (1 μM, 1 h), or a combined treatment of PARGi and PARPi (1 μM, 1 h each). Cells were lysed and chromatin fractions were immunoprecipitated with A-protein magnetic beads. FLAG IP under denaturing conditions was followed by immunoblotting with the indicated antibodies.

(C)Representative images of a PLA for FLAG-tagged TOP3A and endogenous PAR in U2OS cells. PLA with FLAG + PAR antibody either in pretreatment of DMSO (drug solvent), PARGi or PARPi alone (1 μM, 1 h), or combined treatment of PARGi and PARPi (1 μM, 1 h each). DMSO was used in the control samples. Scale bar, 5 μm.

(D) Quantitation of the TOP3A-FLAG and PAR PLA foci shown in (C). Results are represented as SuperPlots³⁵ showing individual data points in different colors, the average from each experiment, and the mean of the averages ± SD. ****p < 0.0001, ordinary one-way ANOVA with Tukey’s multiple comparisons test.

(E) Representative images of a PLA for FLAG-tagged TOP3A and endogenous PARP1 in U2OS cells expressing EV, TOP3A-WT, or TOP3A-R364W. PLA was performed with FLAG + PARP1 antibody in the presence of PARGi (1 μM, 1 h). Scale bar, 5 μm.

(F) Quantitation of the TOP3A-FLAG and PARP1 PLA foci shown in (E). Results are represented as SuperPlots³⁵ showing individual data points in different colors, the average from each experiment, and the mean of the averages ± SD. ****p < 0.0001, ordinary one-way ANOVA with Tukey’s multiple comparison test.

To investigate whether the interaction of FEN1 and TOP3A is cell cycle dependent, we treated TOP3A-R364W-transfected U2OS cells with 10 μM EdU for 40 min to label S-phase cells, followed by EdU click reaction for 1 h, and then performed PLA using anti-FLAG and anti-FEN1 antibodies. PLA foci were observed only in the EdU-positive cells, indicating that the interaction of FEN1 and TOP3A is specific to replicating cells (Figure S2D).

Next, we hypothesized that the induction of TOP3A-DPCs may also activate PARP1, as FEN1 has recently been shown to be recruited by PARylation for the repair of ultraviolet-induced lesions.³⁶ Given the prevalence of TOP3A-DPCs in S-phase cells, lagging strand ssDNA regions behind replication forks are logical hotspots for TOP3A activity,¹¹ as well as for PARylation.^16,17 Thus, we posited that PARylation may recruit FEN1 to TOP3A-DPCs, thereby promoting TOP3A-DPC repair. To address this possibility, we analyzed the effects of talazoparib, a potent PARPi,^20,37 on the interaction between FEN1 and TOP3A. PARP inhibition led to a significant reduction of TOP3A-FEN1 PLA foci in TOP3A-R364W-expressing cells (Figure 2A). Consistent with this observation, immunoprecipitation (IP) of the chromatin fractions of FLAG-tagged TOP3A-R364W-expressing cells with FLAG antibody revealed that endogenous FEN1 interacts with FLAG-tagged TOP3A (Figure 2B).

Because we found that FEN1 interacts with TOP3A in a PARP-dependent manner, we tested whether the TOP3A-DPC and FEN1 interaction is dependent on PAR polymers. Comparing pull-downs of FLAG-tagged TOP3A immunoprecipitates with and without prior incubation with the PARG inhibitor (PARGi)^16,18 showed that PARG inhibition (i.e., hyperPARylation) enriched both FEN1 and PARP1 in the TOP3A immunoprecipitates (compare lanes 3 and 4 in Figure 2B). However, inhibition of FEN1 in PARG-inhibited cells did not affect TOP3A’s interaction with PARP1 (Figure S2E), suggesting that FEN1 is downstream of the TOP3A-PARP1 interaction. As hyperPARylation promotes the FEN1 and TOP3A interaction and PARP inhibition disrupts it (Figures 2A, bottom, and 2B, lane 6), we next tested the interaction between TOP3A and PAR polymers by PLA. The appearance of PLA foci in FLAG-tagged TOP3A-R364W-expressing U2OS cells using anti-FLAG and anti-PAR antibodies, but not in controls either without antibodies or with a single antibody (Figure S2B), confirmed the interaction between TOP3A and PAR polymers (second image in Figure 2C). Consistent with the role of PARylation for the recruitment of FEN1 to TOP3A-DPCs, inhibition of PARP with talazoparib suppressed the PLA foci formed between TOP3A and PAR polymers (Figures 2C and 2D). To specifically examine whether the association of TOP3A and PARP1 is induced by TOP3A-DPCs, we compared the number of FLAG-tagged TOP3A and PARP1 PLA foci formed in TOP3A-R364W-expressing cells with that of mock-transfected (EV) and TOP3A-WT-transfected cells. We confirmed the specificity of the PLA using no antibodies and each single FLAG and PARP1 antibody (Figure S2C). While mock-transfected cells did not form PLA foci, we observed increased PLA foci in TOP3A-R364W-expressing cells in comparison with TOP3A-WT-expressing cells (Figures 2E and 2F), suggesting that TOP3A interacts with PARP1 when it is trapped on chromatin.

PARylation of TOP3A-DPCs induces the removal of TOP3A-DPCs by FEN1

To determine whether cellular TOP3A-DPCs are PARylated, we adapted the modified RADAR assay (Figure 3A), which we previously developed for the detection of ubiquitylated or SUMOylated TOP-DPCs,²⁸ by including an IP step of FLAG-tagged TOP3A. Following TOP3A transfection, RADAR assay samples were subjected to FLAG IP followed by digestion with benzonase to remove the DNA bound to TOP3A-DPCs, electrophoresed in SDS-PAGE, and immunoblotted with an anti-PAR antibody. Using this assay, we confirmed that TOP3A-DPCs were detectable only in TOP3A-R364W-expressing cells (lane 3 in Figure 3B, middle). Upon PARG inhibition, cellular PARylation of TOP3A-DPCs was readily detected in cells transfected with TOP3A-R364W, while no PARylated TOP3A-DPCs were observed in TOP3A-WT- and mock-transfected cells, as they do not form TOP3A-DPCs (lane 6 in Figure 3B, top). This result clearly indicates PARylation of the TOP3A-DPCs in the TOP3A-R364W-expressing cells. Notably, the PARylation signal of TOP3A-DPCs was only seen in the presence of the PARGi (compare lanes 3 and 6 in Figure 3B), indicating that TOP3A-DPC PARylation is dynamic and transient. To confirm the specificity of the modified RADAR assay, the same benzonase-treated samples (as in Figure 3B, top) were subjected to SDS-PAGE and immunoblotting with an anti-TOP3A antibody (Figure 3B, middle), which detected a protein band corresponding to the size of human TOP3A.

Figure 3. PARP1-mediated PARylation of TOP3Accs.

(A) Scheme of the modified RADAR assay including FLAG IP for immunodetection of TOP3A-DPCs and their PARylation using anti-TOP3A and anti-PAR antibody.

(B) HyperPARylation stabilizes TOP3A-DPCs. RADAR assay samples were prepared from mock-transfected (EV) HCT116 cells or HCT116 cells transfected with FLAG-tagged TOP3A-WT or TOP3A-R364W plasmid constructs for 48 h in the absence or presence of PARGi (1 μM, 1 h pretreatment). Equal amounts (20 μg DNA) of RADAR assay samples were subjected to FLAG IP, digested with benzonase nuclease, run on SDS-PAGE, and immunoblotted with anti-PAR and anti-TOP3A antibodies. EV, empty vector transfected. Equal DNA loading was tested with an anti-dsDNA antibody.

(C) PARylation of recombinant TOP3A protein by recombinant PARP1 protein. Following 20 min incubation at room temperature, samples were subjected to IP of TOP3A, followed by western blotting with anti-PAR and anti-TOP3A antibodies.

(D) Following 48 h transfection of FLAG-tagged TOP3A-WT, FLAG-tagged TOP3A-R364W expression, or EV, HCT116 cells were pretreated with 1 μM PARGi for 1 h, followed by chromatin fractionation under denaturing conditions and IP with anti-FLAG antibody. Cells expressing TOP3A-R364W were also pretreated with siRNA against TOP3A for 24 h. The immunoprecipitates and input samples were probed with the indicated antibodies.

(E) Immunoblotting of FLAG immunoprecipitates from U2OS cells expressing TOP3A-R364W in the presence of PARGi (1 μM, 1 h). Cells were lysed under denaturing conditions and chromatin fractions were subjected to IP either in siControl conditions or siRNA against TOP3A (siTOP3A), PARP1 (siPARP1), and PARP-inhibited (1 μM, 1 h) conditions. The immunoprecipitates and input samples were probed with the indicated antibodies.

(F) TOP3A PARylation does not directly affect TOP3A cleavage or reversal activities. Top: Cy3-labeled gapped single-stranded DNA (ssDNA) substrate used for TOP3A activity assay. Bottom: representative gel of TOP3A activity assay. Cleavage sites are indicated by arrows. Reaction products, cleavage, and reversal were generated by TOP3A with the DNA substrate alone, unmodified TOP3A (no PAR), and PARylated TOP3A. Purified TOP3A core (1–637 aa)-RMI1 (1–216 aa) was used in this experiment.

An in vitro PARylation assay using recombinant TOP3A and PARP1 proteins followed by IP of TOP3A revealed that the PARylation signal originated only from the samples with TOP3A (lane 3 in Figure 3C, top), confirming that TOP3A itself is a PARylation substrate (Figure 3C). Notably, the intensity of the TOP3A signal obtained after probing with the anti-TOP3A antibody decreased upon PARylation (compare lanes 1 and 3 in Figure 3C, bottom). In addition, we only observed a minor upper band of TOP3A uponits PARylation (lane 3 in Figure 3C), consistent with the possibility that the TOP3A antibodies fail to recognize PARylated TOP3A.

To further establish that TOP3A-DPCs are directly PARylated in cells, FLAG IP was performed under denaturing conditions to minimize the possibility that PARylation signals could originate from proteins associated with TOP3A after pulling down FLAG. The PARylation signal was seen in the TOP3A-R364W-expressing cells but not in the cells expressing TOP3A-WT and EV, indicating that the PARylation signal originated from TOP3A-DPCs (Figure 3D). Suppression of the TOP3A-DPC PARylation signal by treating the TOP3A-R364W-expressing cells with siRNA against TOP3A (last lane in Figure 3D, top) confirmed the PARylation of TOP3A-DPCs in the presence of the PARGi. Based on recent observations of reversible and rare DNA PARylation of adenine bases in mammalian cells,³⁸ we tested whether the PARylation signal associated with TOP3A-DPCs could emanate from the DNA fragments linked to TOP3A-DPCs in TOP3A-R364W-expressing cells. Proteinase K treatment of the FLAG IP samples from RADAR assays or chromatin fractions eliminated the PAR signal of TOP3A-DPCs (Figures S4A and S4B), ruling out PARylation of DNA bound to TOP3A and establishing the direct PARylation of TOP3A within the TOP3A-DPCs.

To demonstrate that PARP1 catalyzes TOP3A-DPC PARylation, we knocked down PARP1 by siRNA-mediated gene silencing or treated TOP3A-R364W-expressing cells with talazoparib under PARG-inhibited conditions. FLAG-tagged TOP3A IP from chromatin extracts using an anti-FLAG antibody showed that both PARP1 downregulation and inhibition suppressed the PARylation of TOP3A-DPCs (Figure 3E), which demonstrates that TOP3A-DPCs are PARylated by PARP1.

Because PARP1-mediated PARylation of proteins at DNA-damaged sites has been considered to be one of the earliest DDRs, we also measured global PARylation in whole-cell extracts in the presence of the PARGi. TOP3A-R364W-expressing cells displayed increased global cellular PARylation and PARP1 expression in comparison with TOP3A-WT- and mock-transfected cells (Figure S3A), demonstrating that TOP3A-DPCs induce global PARP activation in addition to TOP3A-DPC PARylation.

Next, we tested whether TOP3A PARylation affects the biochemical activity of TOP3A. Using denaturing urea-PAGE gel-based assays with recombinant proteins, we measured TOP3A cleavage activity as the generation of cleavage products and TOP3A reversal activity as the disappearance of TOP3A cleavage products. Purified recombinant TOP3A protein was incubated with a Cy3-labeled DNA gap substrate (Figure 3F, top) for cleavage and reversal experiments with 0.4 M NaCl under PARylation and no-PARylation conditions. DNA substrate and cleavage products were separated by denaturing urea-PAGE gel electrophoresis. The addition of PARP1 and NAD+ had no impact on TOP3A activity, either cleavage or reversal (Figures 3F and S3B), indicating that TOP3A remains active despite its PARylation and that PARylation of TOP3A-DPCs had no detectable effect on their steady-state levels or spontaneous reversal.

Epistasis of FEN1 and PARP1 in TOP3A-DPC repair

To demonstrate that PARylation promotes the excision of TOP3A-DPCs, we treated HCT116 cells transfected with TOP3A-R364W with either the PARPi (talazoparib) or the PARGi (PDD00017273). As shown in Figure 4A, inhibiting PARP caused an elevation in TOP3Accs, while PARG inhibition did not. These data suggest that PARP1-mediated PARylation facilitates the repair of cellular TOP3A-DPCs.

Figure 4. PARP inhibition increases TOP3A-DPCs and is epistatic with FEN1 for TOP3A-DPC repair.

(A) Top: outline of the experimental protocol. TOP3A-R364W-transfected HCT116 cells were pretreated with either DMSO (drug solvent), PARPi (1 μM, 1 h), or PARGi (1 μM, 1 h) before harvest and RADAR assays. Left: representative slot blot probed with anti-TOP3A antibody. Right: quantitation of three independent experiments as shown on the left. Error bar represents mean ± SD. Ordinary one-way ANOVA with Dunnett’s multiple comparison test. DMSO was used as a control condition. **p < 0.01 (control vs. PARPi) and ns: not significant (control vs. PARGi).

(B) Epistasis of FEN1 and PARP1. HCT116 cells were transfected with TOP3A-R364W plasmid constructs for 48 h and co-transfected with either siControl or siFEN1 for 72 h. Before harvest, cells were treated with either the PARPi talazoparib (1 μM, 1h) or DMSO (drug solvent control). TOP3A-DPCs were isolated by RADAR assay and slot blotted with anti-TOP3A antibody. Representative slot blots are shown (left). Right: quantitation of TOP3A-DPCs. Error bars represent data mean ± SD (n = 3 independent experiments). p values were obtained from ordinary one-way ANOVA with Dunnett’s multiple comparison test. *p < 0.05 (siControl vs. siFEN1), **p < 0.01 (siControl vs. PARPi), ***p < 0.001 (siControl vs. siFEN1 + PARPi), and ns: not significant (PARPi vs. siFEN1 + PARPi).

(C) Left: representative slot blot demonstrating the epistasis of FEN1 and PARP1 for TOP3A-DPC repair. HCT116 cells were transfected with TOP3A-R364W for 48 h. Before harvest, cells were treated with either DMSO (drug solvent) or the PARPi talazoparib (1 μM, 1h), FEN1i (10 μM, 1 h), and a combination of PARPi and FENi. Right: quantitation of the 3 independent experiments as shown on the left. Error bars represent mean ± SD. p values were obtained from ordinary one-way ANOVA with Dunnett’s multiple comparison test. **p < 0.01 (control vs. FEN1i), ***p < 0.01 (control vs. PARPi), ***p < 0.01 (control vs. FEN1i + PARPi), and ns: not significant (PARPi vs. FEN1i + PARPi). DMSO was used a control.

(D) Left: representative confocal microscopy images of γH2AX immunostaining of U2OS cells transfected as indicated. Scale bar, 10 μm. DMSO (drug solvent) was used in the control samples. Right: quantification plot showing fluorescence intensities of γH2AX signal per nucleus (analyzed by ImageJ). Data are the mean ± SD (n = 100 cells for each condition). Two-tailed unpaired t test with Welch’s correction. ***p < 0.001 (control vs. FENi), ***p < 0.001 (control vs. PARPi), ***p < 0.001 (control vs. FEN1i + PARPi), and ns: not significant (PARPi vs. FEN1i + PARPi).

(E) Left: representative images of colony formation assays in HCT116 cells transfected with TOP3A-R364W. Right: quantification of the data from experiments as shown on the left. Clonogenic survival histogram data of each condition are presented after normalization with their respective colony formation efficiency in NT condition. Error bars represent data mean ± SD (n = 3 independent experiments). Ordinary one-way ANOVA with Dunnett’s multiple comparison test. ****p < 0.0001 (control vs. FENi), ****p < 0.0001 (control vs. PARPi), ****p < 0.0001 (control vs. FENi + PARPi), and ns: not significant (PARPi vs. FEN1i + PARPi).

(F) Micronucleus frequency histograms under the condition indicated in (E). Error bar represents data mean ± SD (n = 3 independent experiments). p values were determined by ordinary one-way ANOVA with Dunnett’s multiple comparison test. ***p < 0.001 (control vs. FENi), ***p < 0.001 (control vs. PARPi), ***p < 0.001 (control vs. FENi + PARPi), and ns: not significant (PARPi vs. FEN1i + PARPi).

We next investigated whether TOP3A-DPC PARylation is coupled with FEN1-dependent repair. PARP inhibition in FEN1-depleted cells caused no further increase in TOP3A-DPC levels, in comparison with PARP-inhibited and FEN1-deficient cells alone (Figure 4B), indicating an epistasis of FEN1 and PARP1 in TOP3A-DPC repair. We confirmed the epistatic relationship of FEN1 and PARP1 in TOP3A-DPC repair after combined treatment with the PARPi and FEN1i in TOP3A-R364W-expressing HCT116 cells by measuring the level of TOP3A-DPCs (Figure 4C). We also assessed cellular fitness and viability by inhibiting PARP1, FEN1, and dual PARP1/FEN1 after TOP3A-DPC induction by measuring DNA damage (γH2AX immunofluorescence), genome stability (micronuclei frequency), and colony formation efficiency. Dual inhibition of FEN1 and PARP1 caused no further increase in γH2AX and micronuclei (Figures 4D and 4F) and did not further affect cell viability (Figure 4E) in comparison with single FEN1- or PARP1-inhibited conditions, corroborating the epistasis of FEN1 and PARP1 in TOP3A-DPC removal.

PARylation prevents the ubiquitylation of TOP3A-DPCs and their processing by SPRTN

We recently reported the ubiquitylation of TOP3A-DPCs prior to their processing by SPRTN.¹¹ To examine the interplay between TOP3A-DPC PARylation and ubiquitylation, we assessed the impact of PARylation on the ubiquitylation of TOP3A-DPCs. To this end, we performed modified RADAR assays in HCT116 cells expressing FLAG-tagged TOP3A-R364W, including FLAG IP, and probed TOP3A-DPCs with an anti-ubiquitin antibody after pretreatment with either a PARPi or PARGi (Figure 5A). While, as expected,¹¹ ubiquitin-activating enzyme E1 (UBE1) inhibition by TAK243 suppressed the ubiquitylation of TOP3A-DPCs while increasing TOP3A-DPCs (Figure 5A, compare lanes 1 and 2, middle and bottom, respectively), PARP inhibition caused an elevation of ubiquitylated TOP3A-DPCs (Figure 5A, compare lanes 1 and 4, middle). This result suggests that the PARylation suppresses the ubiquitylation of TOP3A-DPCs. TAK243 treatment combined with PARGi further increased the levels of PARylated TOP3A-DPCs and TOP3A-DPCs compared to PARG inhibition alone (Figure 5A, compare lanes 3 and 5, top), implying that ubiquitylation negatively affects PARylation.

Figure 5. PARylation suppresses the ubiquitylation of TOP3A-DPCs and their processing by SPRTN.

(A) Representative modified RADAR assay including FLAG IP using the protocol outlined in Figure 3A and showing that TOP3A-DPC ubiquitylation is enhanced by PARPi and inhibited by the ubiquitin (Ub)-activating enzyme E1 inhibitor (UAE1i). Following 48 h transfection with FLAG-tagged TOP3A-R364W, HCT116 cells were treated with the UAE1i (TAK243; 1 μM, 1 h), PARGi (10 μM, 1 h), PARPi (talazoparib; 1 μM, 1 h), PARGi + UAE1i, or PARPi + UAE1i before harvest and RADAR assay. Equal amounts (20 μg DNA) of RADAR assay samples were subjected to IP with FLAG antibody, digested with benzonase nuclease, run on SDS-PAGE, and immunoblotted to detect TOP3Accs with anti-FLAG antibody. The PARylation and ubiquitylation of TOP3Accs were detected with anti-PAR and anti-Ub antibodies. 4%–12% and 8% SDS-PAGE gels were used for PAR and Ub detection.

(B) PARP inhibition induces SPRTN interaction with TOP3A-DPCs. U2OS cells expressing FLAG-tagged TOP3A-R364W after 48 h transfection were pretreated with either DMSO (drug solvent), the PARPi talazoparib (1 μM, 1 h), or UAE1i (TAK243; 1 μM, 1 h) alone and a combination of them (PARPi + TAK243) before cell fixation and PLA. Representative images of a PLA for FLAG-tagged TOP3A and endogenous SPRTN. Scale bars, 5 μm. DMSO was used in the control samples.

(C) Quantitation of the PLA foci shown in (B). Results are represented as SuperPlots³⁵ showing individual data points in different colors, the average from each experiment, and the mean of the averages ± SD. p values were obtained by ordinary one-way ANOVA with Tukey’s multiple comparisons test. ****p < 0.0001 (control vs. PARPi), ****p < 0.0001 (control vs. TAK243), and ****p < 0.0001 (PARPi vs. PARPi + TAK243).

(D) FEN1 inhibition induces SPRTN interaction with TOP3A-DPCs. U2OS cells transfected with FLAG-tagged TOP3A-R364W were pretreated with either DMSO (drug solvent) or FEN1i (10 μM, 1 h) before cell fixation and PLA. Representative images of a PLA for FLAG-tagged TOP3A and endogenous SPRTN. Scale bars, 5 μm. DMSO was used in the control samples.

(E) Quantitation of PLA foci shown in (D). Results are represented as SuperPlots³⁵ as described in (C). p values were determined by two-tailed unpaired t test with Welch’s correction. ****p < 0.0001.

Recent studies have shown that ubiquitylation of DPCs is required for their repair by SPRTN²⁹ and that SPRTN is involved in TOP3A-DPC repair.¹¹ To examine whether PARylation modulates the recruitment of SPRTN to TOP3A-DPCs, we treated cells expressing TOP3A-R364W with talazoparib for 1 h before fixation and PLA. PARP inhibition increased PLA foci between SPRTN and FLAG-tagged TOP3A (Figures 5B, 5C, and S5A). The increased recruitment of SPRTN at TOP3A-DPC sites in chromatin upon PARP inhibition suggests that PARylation suppresses the recruitment of SPRTN to TOP3A-DPCs. In contrast, suppression of PLA foci between SPRTN and FLAG-TOP3A by ubiquitylation inhibition (TAK243 treatment) indicates the importance of ubiquitylation for the recruitment of SPRTN at TOP3A-DPC sites (Figures 5B and 5C). We further assessed SPRTN recruitment after FEN1 inhibition in TOP3A-R364W-expressing cells. Like PARP inhibition, FEN1 inhibition increased the SPRTN-TOP3A interaction (Figures 5D and 5E). These results suggest that a failure to repair TOP3A-DPCs via PARP1-FEN1 channels repair to the ubiquitin and SPRTN pathway.

Coordination of the FEN1-PARylation and SPRTN-ubiquitylation pathways for TOP3A-DPC repair

Ubiquitin signaling has been linked to both the proteasome- and SPRTN-mediated repair of TOP1-DPCs.^3,28,39 To examine the effect of ubiquitin on TOP3A-DPC repair, we performed RADAR assays in TOP3A-R364W-expressing cells in the presence of TAK243 and the proteasome inhibitor bortezomib. While proteasome inhibition did not change TOP3A-DPC levels as reported previously,¹¹ inhibition of ubiquitin-activating enzyme E1 (UAE1) by TAK243 caused an almost 2-fold elevation of TOP3A-DPCs (Figure 6A). To define the relationship between ubiquitylation and SPRTN, we pretreated SPRTN-deficient TOP3A-R364W-expressing cells with TAK243 before cell harvest and RADAR assay. UAE1 inhibition by TAK243 treatment in SPRTN-depleted cells caused no further increase in TOP3A-DPC levels in comparison with UAE1-inhibited and SPRTN-deficient cells alone (Figure 6B), indicating an epistasis of UAE1 inhibition and SPRTN depletion for the repair of TOP3A-DPCs.

Figure 6. Ub signaling and SPRTN, and SPRTN- or FEN1-mediated TOP3A-DPC repair.

(A) Representative slot blots of TOP3A-DPCs detected using anti-TOP3A antibody (top). HCT116 cells were transfected with TOP3A-R364W for 48 h. Before harvest, cells were treated with either DMSO (drug solvent), the UAE1i TAK243 (1 μM, 2 h), or the proteasome inhibitor bortezomib (1 μM, 2 h). The bottom shows the quantification of TOP3A-DPCs from 3 independent experiments. Error bars indicate mean ± SD. DMSO was used as a control. p values were obtained by ordinary one-way ANOVA with Dunnett’s multiple comparison test. **p < 0.01 (control vs. TAK243) and ns: not significant (control vs. BTZ).

(B) Epistasis of Ub signaling and SPRTN. HCT116 cells were transfected with either siControl or siFEN1 for 72 and 24 h after co-transfection with TOP3A-R364W for 48 h before performing RADAR assays for TOP3A-DPC detection. Before harvest, cells were treated with either DMSO (drug solvent control) or the UAEi TAK243 (1 μM, 1 h). A representative slot blot is shown (top) with quantitation of TOP3A-DPCs (bottom) from three independent experiments. Error bars indicate mean ± SD. p values were obtained by ordinary one-way ANOVA with Dunnett’s multiple comparison test. ****p < 0.001 (siControl vs. siSPRTN), **p < 0.01 (siControl vs. siControl + TAK243), ****p < 0.0001 (siControl vs. siSPRTN + TAK243), and ns: not significant (siSPRTN vs. siSPRTN + TAK243).

(C) Additive effects of FEN1 and SPRTN depletion on TOP3A-DPCs. HCT116 cells were transfected with siFEN1, siSPRTN, and a combination of siFEN1 and siSPRTN for 72 h and, 24 h later, co-transfected with TOP3A-R364W for 48 h before performing RADAR assays. A representative slot blot is shown (top) with quantitation of TOP3A-DPCs from 3 independent experiments (bottom). Error bars indicate mean ± SD. p values were obtained by ordinary one-way ANOVA with Dunnett’s multiple comparison test. *p < 0.05 (siControl vs. siFEN1), ****p < 0.0001 (siControl vs. siSPRTN), ****p < 0.0001 (siControl vs. siSPRTN/siFEN1), and ****p < 0.0001 (siSPRTN vs. siSPRTN/siFEN1).

(D) Top: representative slot blot showing the additive effects of FEN1 inhibition and SPRTN depletion on TOP3A-DPCs. HCT116 cells were transfected with siSPRTN for 72 h or siControl. Before harvest, cells were treated with either FEN1i (10 μM, 1 h) or DMSO (solvent control), followed by RADAR assay to measure TOP3A-DPCs by slot blot. Bottom: quantitation of 3 independent experiments as shown on the left. Error bars indicate mean ± SD. p values were obtained by ordinary one-way ANOVA with Dunnett’s multiple comparisons test. **p < 0.01 (siControl vs. siSPRTN), **p < 0.01 (siControl vs. siControl + FEN1i), ***p < 0.001 (siControl vs. siSPRTN + FEN1i), and ***p < 0.001 (siSPRTN vs. siSPRTN + FEN1i).

(E) Left: representative images of γH2AX immunostaining of TOP3A-R364W-transfected U2OS cells as indicated. Scale bar, 10 μm. Right: quantification plot showing fluorescence intensities of γH2AX signal per nucleus (analyzed by ImageJ). Data are the mean ± SD (n = at least 100 cells for each condition). Two-tailed unpaired t test with Welch’s correction. ****p < 0.001 (siControl vs. siFEN1), ****p < 0.0001 (control vs. siSPRTN), ****p < 0.0001 (siControl vs. siFEN1/siSPRTN), and ****p < 0.0001 (siSPRTN vs. siFEN1/siSPRTN).

(F) Top: representative images of colony formation assays in indicated conditions of HCT116 cells after transfection with TOP3A-R364W. Right: quantification of the data from experiments as shown on the left. Clonogenic survival histogram data of each condition are presented after being normalized with their respective colony formation efficiency in NT conditions. Error bars represent mean ± SD (n = 3 independent experiments). Ordinary one-way ANOVA with Dunnett’s multiple comparisons test. ****p < 0.001 (siControl vs. siFEN1), ****p < 0.001 (control vs. siSPRTN), ****p < 0.0001 (siControl vs. siFEN1/siSPRTN), and **p < 0.01 (siSPRTN vs. siFEN1/siSPRTN).

Having demonstrated that both ubiquitylation and PARylation are important for TOP3A-DPC repair, we examined the potential redundancy between FEN1 and SPRTN. Concurrent depletion of both FEN1 and SPRTN by siRNA (Figure S6A) enhanced TOP3A-DPCs to higher levels than depletion of FEN1 or SPRTN individually, indicating an additive effect of FEN1 and SPRTN for TOP3A-DPC repair (Figure 6C). Consistent with this conclusion, the increased level of TOP3A-DPCs upon pharmacological inhibition of FEN1 showed additive effects of FEN1 and SPRTN in TOP3A-DPC removal in SPRTN-depleted cells (Figure 6D). We further assessed cellular fitness and viability by silencing SPRTN or FEN1 alone and silencing both SPRTN and FEN1 in cells forming TOP3A-DPCs. Disruption of both FEN1 and SPRTN further increased γH2AX and micronuclei levels (Figures 6E and S6B) and decreased cell viability (Figure 6F) in comparison with single SPRTN- or FEN1-silenced conditions. Based on these results, we propose the existence of two parallel pathways for the removal of TOP3A-DPCs: one comprising FEN1 and PARylation and the other SPRTN and ubiquitylation.

Coordination of the FEN1 and MRE11/ATM pathways for TOP3A-DPC repair

To explore the coordination between FEN1-mediated and the recently reported MRE11/ATM-mediated¹¹ TOP3A-DPC repair, we examined the levels of TOP3A-DPCs in FEN1/MRE11 double-deficient cells by RADAR assay. MRE11 depletion in FEN1-deficient cells caused a further increase in TOP3A-DPC levels in comparison with MRE11 or FEN1 knockdown alone (Figure 7A), indicating an additive effect of MRE11 and FEN1 in TOP3A-DPC repair. In addition, ATM inhibition in FEN1-depleted cells caused a further increase in TOP3A-DPC levels in comparison with ATM-inhibited and FEN1-deficient cells alone, also indicating an additive relationship between FEN1 and ATM (Figure 7B). To further examine the interplay and specificity of the FEN1- and ATM/MRE11-mediated TOP3A-DPC repair pathways, we examined the effect of FEN1 depletion and ATM inhibition or MRE11 depletion on TOP3A-DPC accumulation in S and G2/M phases in cell-cycle-synchronized¹¹ cells. The activity of FEN1 in removing TOP3A-DPCs was observed in S-phase cells but not in G2/M-phase cells (Figure 7C). In contrast, the effects of MRE11 and ATM on TOP3A-DPCs were mainly observed in G2/M-phase cells, consistent with our prior findings.¹¹

Figure 7. Additive relationship between FEN1 and MRE11/ATM, and proposed model of parallel pathways for TOP3A-DPC repair driven by their post-translational modifications.

(A) Additive effects of FEN1 and MRE11 depletion on TOP3A-DPCs. U2OS cells were transfected with siFEN1, siMRE11, and a combination of siFEN1 and siMRE11 for 72 h and, 24 h later, co-transfected with TOP3A-R364W for 48 h before performing RADAR assays. A representative slot blot is shown (top) with quantitation of TOP3A-DPCs from 3 independent experiments (bottom). Error bars indicate mean ± SD. p values were obtained by ordinary one-way ANOVA with Dunnett’s multiple comparisons test. **p < 0.01 (siControl vs. siFEN1), ***p < 0.001 (siControl vs. siMRE11), ****p < 0.0001 (siControl vs. siFEN1/siMRE11), and ***p < 0.001 (siMRE11 vs. siFEN1/siMRE11).

(B) Additive effects of FEN1 depletion and ATM inhibition on TOP3A-DPCs. U2OS cells were transfected with TOP3A-R364W plasmid for 48 h. Cells were co-transfected with either siControl or siFEN1 for 72 h. Before harvest and RADAR assays, cells were treated with either DMSO or the ATMi KU-55933 (20 μM) for 2 h. A representative slot blot is shown (top) with quantitation of TOP3A-DPCs from 3 independent experiments (bottom). Error bars indicate mean ± SD. p values were obtained by ordinary one-way ANOVA with Dunnett’s multiple comparisons test. **p < 0.01 (siControl vs. siFEN1), **p < 0.01 (siControl vs. ATMi), ***p < 0.001 (siControl vs. siFEN1 + ATMi), and **p < 0.01 (siFEN1 vs. siFEN1 + ATMi).

(C) Differential roles of FEN1 and MRE11/ATM for TOP3A-DPC repair as a function of cell cycle. U2OS cells were transfected with either siControl or siFEN1, siMRE11 followed by transfection with TOP3A-R364W. siControl cells were treated with the ATMi KU-55933 (20 μM) for 2 h before harvest. Cells (in S and G2/M phases) were collected after synchronization by double-thymidine block. RADAR assays were performed with anti-TOP3A antibody. Representative slot blot images are shown (top). Quantitation from 3 independent RADAR assays is shown on the bottom. Error bars indicate mean ± SD. p values were obtained by two-way ANOVA with Tukey’s multiple comparisons test. ****p < 0.0001 and ns: not significant.

(D) Two types of post-translational modifications, PARylation and ubiquitylation, direct the processing of trapped TOP3Accs (TOP3A-DPCs). Left: PARylation of TOP3A-DPCs recruits FEN1. Right: following dePARylation by PARG, ubiquitylation recruits SPRTN for proteolytic degradation of TOP3A-DPCs.

DISCUSSION

The present study reveals the regulatory mechanisms for TOP3A-DPC repair linking FEN1 and PARP1. Using cellular and biochemical assays, we show that PARP1 transiently PARylates TOP3A without blocking its catalytic activity (Figure 3). PARylation of TOP3A-DPCs recruits FEN1 to remove TOP3A-DPCs (Figure 2). Genetic evidence suggests that PARP1 and FEN1 are epistatic for the repair of TOP3A-DPCs (Figure 4). Transient PARylation suppresses the ubiquitylation of TOP3A-DPCs and their recognition by SPRTN, and vice versa (Figure 5). We propose that two coordinated repair pathways remove TOP3A-DPCs: PARylation-driven FEN1 and ubiquitylation-driven SPRTN (Figures 6 and 7), and that PARP1 plays a key role in regulating the repair of TOP3A-DPCs (as summarized in Figure 7D).

FEN1 catalytic activity is important for DNA replication and repair,⁴⁰ and here, we show that blocking FEN1 endonuclease activity genetically and pharmacologically³² results in a failure to remove the TOP3A-DPCs induced by our TOP3A-poisoning mutant (TOP3A-R364W)¹¹ (Figure 1). Additionally, we found detectable levels of endogenous TOP3A-DPCs in FEN1 knockdown cells (Figure 1B), suggesting that FEN1 repairs abortive TOP3A-DPCs in baseline conditions. The decreased TOP3A-DPC levels after ectopic co-expression of FEN1-WT and TOP3A-R364W constructs in cells with equal amounts of each plasmid (1:1) suggest that FEN1 is a rate-limiting factor for TOP3A-DPC repair. Our finding that FEN1, a member of the Rad2 nuclease family, is required for TOP3A-DPC repair in human cells generalizes the earlier findings in yeast that the structure-specific nuclease Rad2 (XPG in human) can remove TOP2-DPCs as an alternative repair pathway.⁴¹

Beyond the role of PARP1²³,⁴² and TOP3A^11,26 in DNA replication, we demonstrate the importance of FEN1, TOP3A, and PARP1 for TOP3A-DPC repair, where PARP1-mediated PARylation of TOP3A-DPCs recruits FEN1. Protein-protein interactions are critical in guiding FEN1 into multiple cellular metabolic pathways, which is likely due to the fact that FEN1 interacts with different proteins including PCNA, RPA, APE1, PARP, DNA polymerases, and P300 to form protein complexes that play specific roles in DNA replication and repair.⁴³ The DNA-protein interactions modes of FEN1 also dictate FEN1 activity, as FEN1 binds to the ssDNA region of gapped DNA substrates.⁴⁴ Biochemically, FEN1 activity on gapped-fork and gapped-duplex substrates is higher than for the other members of RAD2 family, such as EXO1 and GEN1. Similarly, TOP3A-DPCs may form in the ssDNA regions behind replication forks, considering the known activity of FEN1 on such substrates during replication.^11,13 Consistent with this possibility, we find an interaction of TOP3A with FEN1 in TOP3A-R344W-expressing cells that form TOP3A-DPCs (Figures 2B and 2C). As evidenced by chromatin IP and PLAs (Figures 2B and 2C), we find that this interaction is increased by stable hyperPARylation in cells treated with a PARGi and is suppressed by PARP inhibition. Our findings of PARP1’s interaction with TOP3A in TOP3A-R364W-expressing cells is consistent with the notion that PARP1 is the major producer of cellular PAR in response to DNA damage^21,23 (Figures 2B and 2E).

Transient PARylation of TOP3A-DPCs (Figure 3B) and pulled down TOP3A-R364W from chromatin fraction under denaturing conditions (Figure 3D) indicate the direct PARylation of TOP3A in TOP3A-DPCs (Figure 3D). An interplay between PARylation and the DNA repair enzyme has already been established.^22,23 A prominent example is that S-phase PARP1 activity recruits the DNA repair protein XRCC1 and thereby facilitates SSB repair and/or base excision repair (BER).¹⁶ Besides BER, PARP1 also plays a role in nucleotide excision repair (NER) by PARylating the NER proteins xeroderma pigmentosum complementation group A (XPA) and XPC.^45–47 The reduction of PLA foci between FEN1 and TOP3A upon PARP inhibition is consistent with PARylation acting as a signal for FEN1 recruitment at TOP3A-DPC sites (Figure 2A). Accordingly, an increased accumulation of TOP3A-DPCs was observed in TOP3A-R364W-expressing cells after PARPi treatment (Figure 4A). The epistasis of PARP1 and FEN1 further supports the requirement of PAR signaling in FEN1-mediated TOP3A-DPC repair (Figures 4B–4F).

Ubiquitylation of TOP-DPCs is known to initiate their proteasomal degradation and SPRTN-dependent proteolysis.^1,3 The present study shows that cellular TOP3A-DPCs undergo at least two post-translational modifications: PARylation and ubiquitylation (Figures 3 and 5A). Although the proteasome is dispensable for TOP3A-DPC repair,¹¹ our current findings suggest that ubiquitylation of TOP3A-DPCs recruits SPRTN (Figure 5B) and that the ubiquitin binding UBZ domain of SPRTN regulates this recruitment. ^11,29,48,49 This possibility is supported by our observations that the ubiquitin-binding domain of SPRTN is required for the proteolytic processing of cellular TOP3A-DPCs.¹¹ In the current work, we provide evidence that the ubiquitination of TOP3A-DPCs is coordinated with their PARylation (Figure 5). While PARylation of TOP3A-DPCs enhances FEN1 recruitment, we find that it also suppresses the ubiquitylation of TOP3A-DPCs (Figure 5A). Notably, while PARP inhibition drives ubiquitylation of TOP3A-DPCs and sequential SPRTN recruitment to recognize them, stabilization of PARylated TOP3A-DPCs by PARG inhibition has no impact on the level of TOP3A-DPCs and their ubiquitylation (Figures 5A and 4A). This suggests that only transient hyperPARylation of TOP3A-DPCs suppresses their ubiquitylation. In fact, we find that suppressing ubiquitylation increases TOP3A-DPC PARylation in the presence of PARGi (Figure 5A), which implies that ubiquitylation affects PARylation, and vice versa. These crosstalks between the PARylation and ubiquitylation pathways are an extension of our previous findings for TOP1-DPCs.¹⁸ Like our previous work, the present study also suggests that the cell cycle is a key determinant for pathway choice for TOP3A-DPC repair, with the FEN1-PARP1 pathway prominent in S phase and the MRE11/ATM pathway acting in G2/M phase (Figures 7A–7C).

In summary, our study reveals the PARP1-driven FEN1 pathway acting in coordination with the ubiquitin-mediated SPRTN pathway¹¹ (Figures 6C–6F). We propose two important functions of PARP1 in TOP3A-DPC repair: (1) PARP1-mediated facilitation of FEN1 nuclease-mediated excision of TOP3A-DPCs and (2) PARP1-mediated downregulation of TOP3A-DPC ubiquitylation limiting SPRTN-mediated proteolysis of TOP3A-DPCs (Figure 7D).

Limitations of the study

A limitation is the use of TOP3A self-trapping mutant (TOP3A-R364W) overexpression to generate TOP3A-DPCs and study their repair. We had to rely on our TOP3A-R364W mutant¹¹ because there is no reported drug to poison TOP3A and induce TOP3A-DPCs. While this approach may be advantageous for studying the pathological impact of TOP3A overexpression, it may not necessarily reflect the physiological situation of low endogenous TOP3Accs. Although we demonstrate crosstalk between the PARylation and ubiquitylation of TOP3A-DPCs, our study warrants further investigations to pinpoint further detailed molecular mechanisms and particularly the identification of the E3 ubiquitin ligase(s) involved in the ubiquitylation and subsequent degradation of TOP3A-DPCs. Further studies are also warranted to identify the TOP3A PARylation sites.

STAR★METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Yves Pommier (pommier@nih.gov)

Materials availability

All reagents used and/or generated for this study will be made available from lead contact with Materials Transfer Agreement completion.

Data and code availability

• All data are available in the main text and supplemental information. Uncropped original imaging data supporting the current study have been deposited at Mendeley Data (https://doi.org/10.17632/vgd8fb98vr.1) and are publicly available as of the date of publication.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell lines

Human MCF7, HCT116 and U2OS cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Cat# 084564, Gibco, US) supplemented with fetal bovine serum (10%, Gibco, US), penicillin (100 U/ml), and streptomycin (100 μg/mL, ThermoFischer, US), and maintained at 37°C under a humidified atmosphere and CO₂ (5%).

METHOD DETAILS

Plasmids and transfection

Human FEN1-Myc-FLAG cDNA ORF (CAT#: RC206534) clone was purchased from OriGene.

Site-directed mutagenesis was performed using QuikChange II XL site-directed mutagenesis kit (Agilent Technologies) following the manufacturer’s protocol, and mutations were confirmed by sequencing. D181A-FEN1-Myc-FLAG was generated using oligonucleotides:5′-CGAGGACATGGCCTGCCTCAC.

−3′ and 5′- GTGAGGCAGGCCATGTCCTCG −3′. Human TOP3A plasmids, FLAG-TOP3A-WT and FLAG-TOP3A-R364W were described previously.¹¹ Transfection of expression plasmids was carried out using Lipofectamine 3000 reagents (CAT#: L3000015, ThermoFischer, US) according to the manufacturer’s protocol for 48 h or otherwise indicated in the figure legend.

Generation of small hairpin RNA (shRNA)-mediated gene silencing of FEN1 MCF-7

shRNA oligonucleotides targeting human FEN1, using following primers were cloned into pLKO.1 (Cat#8453, Addgene) digested with EcoRI and AgeI. pLKO.1-shFEN1 or pLKO.1 control vector was simultaneously transfected into LentiX293T (Cat# 632180 Clontech, Japan) with packaging plasmid, pSPAX2 (Cat#12260, Addgene) and envelop plasmid, pMD2.G (Cat#12259, Addgene). After harvesting the medium (3 mL) containing lentiviral particles, we enriched the lentiviral particles by Lent-X Concentrator (Cat# 631231, TaKaRa, Japan) according to the manufacture protocol. The supernatant containing virus was mixed with wild-type MCF-7 cells. The infected cells (Puromycin resistant) were enriched by puromycin drug selection for 72 h. Downregulation of FEN1 expression was confirmed by western blotting using anti-FEN1 antibody.

siRNA transfections

siRNAs were obtained from Horizon Discovery (Dharmacon) and transfected using Lipofectamine RNAiMAX (Invitrogen) according to manufacturer’s protocol. The non-targeting siRNA (siControl) was obtained from Horizon Discovery (Dharmacon) and used as the control. Cells were plated for assays either 48 or 72 h later as indicated in the figure legends. The following siRNA were used: Control siRNA (Control Pool, D-001206-13-5); FEN1 siRNA (SMARTPool, L-010344-00-0005); TOP3A siRNA (ORF: Cat# L-005279-00-0005); MRE11 siRNA (Cat# L-009271-00-0005) and SPRTN siRNA (SMARTPool, L-015442-02-0005).

RADAR assays

Either mock- or expression plasmids-transfected cells (~1 × 10⁶) were washed with PBS and lysed by adding 1 mL DNAzol (ThermoFisher Scientific, Cat#:10503027). Nucleic acids were precipitated following addition of 0.5 mL of 100% ethanol, incubation at −20°C for 5 min and centrifugation (12,000 x g for 10 min). Precipitates were washed twice in 75% ethanol, resuspended in 200 μL TE buffer, heated at 65°C for 15 min, followed by shearing with sonication (40% power for 10 s pulse and 30 s rest 5 times). Samples were centrifuged at 21,000 x g for 5 min and the supernatant containing nucleic acids with covalently bound proteins were collected. Nucleic acid containing protein adducts were quantitated, slot-blotted and TOP3A-DPCs were detected with rabbit polyclonal anti-TOP3A antibody (dilution 1:1000, Proteintech, Rosemont, IL, CAT#: 14525-1-AP).

Proximity ligation assay (PLA)

Proximity ligation assays were performed using the Duolink kit from Sigma Aldrich (Cat# DUO92101) according to manufacturer’s protocol. Briefly, 48 h after TOP3A plasmids transfection, cells were seeded and grown on coverslips in 24 well format. After overnight, cells were treated with either PARPi or PARGi before fixation. After fixation for 15 min at 4°C in 4% paraformaldehyde in PBS cells were permeabilized with 0.25% Triton X-100 in PBS for 15 min at 4°C. The coverslips were blocked with Duolink blocking solution. Primary antibodies were then added at a 1:500 dilution in Duolink antibody diluent and incubated for 1–2 h at room temperature. Proximity ligation assay (PLA) minus and plus probes were diluted 1:5 in the provided dilution buffer, 50 μL of the probe reaction was added to each coverslip and incubated for 1 h at 37°C; the coverslips were then washed twice with buffer A. The provided ligation buffer was diluted 1:5 in water, and then, the ligase was added at a 1:30 dilution; followed by incubation at 37°C for 30 min before washing twice with wash buffer A. The provided amplification buffer was diluted 1:5 in water before adding the provided polymerase at a 1:80 ratio, the amplification reaction was left at 37°C for 100 min, the reaction was quenched by washing twice with buffer B. The coverslips were mounted on slides with DAPI containing mounting medium. Images were captured with a Zeiss LSM 880 Airyscan confocal/super resolution microscope with 63x objective lens. Images were analyzed by ImageJ (Fiji).

EdU labeling followed by PLA

EdU labeling was performed as described⁵⁰ with little modification. Cells were first incubated with 10 μM EdU for 40 min. After washing with PBS once, cells were fixed with 4% paraformaldehyde for 10 min at RT then permeabilized with 0.3% Triton X-100 in PBS, for 3 min on ice. After EdU pulse, cells were blocked with BSA 1% in PBS for 10 min at RT while the Click-iT reaction cocktail was prepared to conjugate the EdU with Alexa Fluor 488 azide (for S phase cells) following the manufacturer’s instructions (Thermo Fisher Scientific, #C10269). The reaction cocktail was freshly prepared before use. Coverslips were incubated with this cocktail for 30–60 min at RT, protected from light in a humidified chamber. Following Alexa Fluor 488 azide conjugation, coverslips were washed twice with PBS before proceeding to the PLA assay as described above.

Denaturing FLAG immunoprecipitation (FLAG-IP)

Either mock-or FLAG-tagged TOP3A plasmids-transfected cells pellets were incubated on ice for 15 min in pre-extraction buffer (25 mM HEPES, pH 7.4, 50 mM NaCl, 1 mM EDTA, 3 mM MgCl₂, 300 mM sucrose, 0.5% Triton X-100), supplemented with protease and phosphatase inhibitors. After centrifugation (5000 g, 5 min) and removal of the supernatant, chromatin pellets were resuspended in modified RIPA buffer containing 0.5% SDS and 300 mM NaCl (10 mM Tris-HCl pH7.5, 300 mM NaCl, 5 mM EDTA, 0.5% SDS, 1% Triton X-100, 1% sodium deoxycholate) supplemented with protease and phosphatase inhibitors. Lysates were homogenized with sonication and incubated with benzonase for 1 h on a rotator at 4°C. After centrifugation (18 000 × g, 10 min), the supernatant was collected and used for protein concentration measurement. 300 μg of the extract were resuspended in 200 μL modified RIPA buffer supplemented with protease and phosphatase inhibitors. Samples were then incubated with 2 μg of anti-FLAG M2 antibody (Sigma Aldrich) on a rotator overnight at 4°C. Protein A/G agarose beads were washed three times for 5 min with RIPA buffer and incubated with the samples for 4 h on a rotator at 4°C. After three washes, proteins were eluted with 2X SDS loading buffer and incubated for 10 min at 95°C on a thermomixer. After centrifugation at 18 000 g (5 min), supernatants were transferred to new tubes, and the proteins separated by SDS-gel electrophoresis. Proteinase K digestion (100 μg/mL final concentration) was performed on FLAG-IP beads samples for 3 h at 37°C. After adding 2X SDS loading buffer, digested samples were incubated for 5 min at 95°C and centrifuged before SDS-PAGE.

Modified RADAR assay (FLAG IP of RADAR samples for detection of PARylated and ubiquitylated TOP3A-DPCs)

For detection of ubiquitylated and PARylated TOP3Accs, we performed DUST assays as described previously.^28,51 First, we performed RADAR assays as described above in either mock- or FLAG-tagged TOP3A-WT and TOP3A-R364W transfected cells (1 × 10⁶). DEPC-treated water resuspended RADAR assay samples containing nucleic acids with covalently bound proteins were then collected, followed by treatment with RNaseA (100 μg/mL) for 1 h at 4°C and addition of 1:10 volume of 3M sodium acetate and 2.5 volume of volume 100% ethanol. Samples were then centrifuged at 21,000 x g for 5 min and the pellets containing nucleic acids and covalent protein-nucleic acid adducts were recovered and further resuspended in DEPC treated water, followed by quantitation of DNA concentration using NanoDrop. Twenty μg of each RADAR assay samples were diluted in IP buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.2% Triton X-100, 5% glycerol, 1 mM DTT, 20 mM N-ethylmaleimide and protease inhibitor cocktail) containing anti-FLAG (M2) antibody and rotated overnight at 4°C. Next day, Protein A/G agarose beads was added and rotated with the samples for another 4 h. Immunoprecipitated samples were washed with IP buffer twice, resuspended in TE buffer. Samples were incubated in then digested with 250 units benzonase nuclease (EMD Millipore, 100 units/μl) by incubation in Thermomixer (1200 rpm, 37°C for 1 h), resuspended in tris-glycine SDS sample buffer (Novex, LC2676), followed by SDS-PAGE electrophoresis for immunodetection of total TOP3A-DPCs and ubiquitylated and PARylated TOP3A-DPCs by probing with rabbit anti-TOP3A (dilution 1:1000, Proteintech, Rosemont, IL, Cat#: 14525-1-AP), mouse anti-ubiquitin antibody (dilution 1:500, Cell Signaling Technology, Cat# 3936), and mouse anti-PAR antibody (dilution 1:1000, Cat # 4335-MC-100), Clone # 10HA), respectively. Co-digestion with Proteinase K (100 μg/mL final concentration) and benzonase was performed on FLAG-IP beads samples from RADAR assays for 3 h at 37°C. After adding 2X SDS loading buffer, digested samples were incubated for 5 min at 95°C and centrifuged before SDS-PAGE.

In-vitro PARylation assay followed by TOP3A IP

Recombinant TOP3A (5 μg) was incubated with 200 nM recombinant PARP1 enzyme (5 μg) in 1x PARylation buffer (50 mM Tris-HCl pH 8.0, 50 mM NaCl, 10 mM MgCl₂, 2% glycerol, 1 mM DTT). 1 mM NAD⁺ were added to the reaction as indicated. The reactions were incubated at room temperature for 20 min, followed by IP as described above using anti-TOP3A antibody and inactivated or by adding SDS sample buffer for Western blotting analysis.

TOP3A activity assay

Oligonucleotides were purchased from IDT. A 37-nt gapped DNA substrate was prepared by annealing Cy3 labeled 5′-GACGTCTATGCGATAGACAGATATTCAGATCATGTCA-3′ DNA oligos with 3′-CTGCAGATACGCT-5′, and 3′-AGTCTAGTACAGT-5′ in 10 mM Tris-HCl, pH7.5, 50 mM NaCl and 1 mM EDTA. The oligo was annealed after heating at 95 °C for 5 min. The DNA substrate was mixed with recombinant human TOP3A (unmodified and PARylated) in 20 μL reaction buffer containing 20 mM Tris-HCl (pH 7.0), 100 mM potassium glutamate, 1 mM MnCl₂, and 2 mM DTT at final concentrations of labeled DNA oligo and TOP3A were 50 nM and 200 nM, respectively. Recombinant TOP3Acore (1-637AA)-Rmi1(1-216AA) was modified by in vitro PARylation assay as described above. After incubated at room temperature for 25 min, cleavage products (20 μL) were mixed with 20 μL 2X formamide gel-loading buffer (10 mM EDTA, 0.025% bromophenol blue, 0.025% Xylene cyanol FF and 0.2% SDS dissolved in formamide), heat denatured at 95°C for 3 min, and separated on a 18% acrylamide gel containing 7 M Urea. Gel was then imaged with Amersham Typhoon laser scanner platform.

To observe reversal of TOP3A cleavage product, 2.2 μL of 10X reversal buffer containing 5 M NaCl was added to the 20 μL reaction sample and incubated for indicated times before mixing with 2X Formamide gel-loading buffer.

Colony formation assay

siControl or gene-silenced and drug-treated HCT116 cells were transfected with R364W-TOP3A-R364W constructs. 48 h after transfection, cells were harvested and serially diluted with fresh medium and plated in 6-well plates (200–300 cells/well) in triplicate. After 12 days, colonies were fixed and stained with 0.5% crystal violet in 20% methanol, and colonies were counted using ImageJ. Data were normalized to the number of seeding cells at Day 0 of each condition.

Alkaline comet assay

As described previously,⁵² alkaline comet assays were performed to quantify single-stranded DNA breaks (SSBs). In brief, shControl and shFEN1 MCF-7 cells after 24 h transfection of either TOP3A-R364W plasmid or mock-transfected (empty vector) were collected for alkaline comet assays using the CometAssay Kit (R&D Systems, Catalog # 4250-050-K) following manufacturer’s instructions. Images were captured using micropscopy with 20X magnification (Zeiss) and tail moment was calculated using OpenComet, a plugin for ImageJ.

MICRONUCLEI ASSESSMENT

TOP3A-R364W-transfected U2OS cells with either control conditions (siControl, DMSO-treated) or gene silenced/drug-inhibited conditions were seeded on chamber slides. After fixation with 4% PFA followed by permealization and blocking step, cells were stained with mounting medium containing DAPI. DAPI stained micronuclei adjacent to main nuclei were imaged with Nikon SoRa super-resolution spinning-disk microscopy and assessed visually. The percentage of the frequency of micronuclei was calculated manually.

QUANTIFICATION AND STATISTICAL ANALYSIS

Western blots and PLA foci were quantified using Image Lab 6.1 and ImageJ (Fiji), respectively and statistical analyses were carried out using GraphPad prism 10 software. Data are provided as means ± standard deviations (SD) from the number of independent experiments performed indicated in each figure legend. Test methods are described in each figure legend. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001 was considered significant and ns: not significant.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Monoclonal ANTI-FLAG® M2 Mouse Monoclonal Antibody	Millipore Sigma	Cat# F1804; RRID: AB_262044
Anti-GAPDH Rabbit Monoclonal Antibody, Unconjugated, Clone 14C10	Cell Signaling Technology	Cat# 2118; RRID: AB_561053
Sheep Anti-Mouse IgG ECL Antibody, HRP Conjugated, GE Healthcare	GE Healthcare	Cat# NA9310-1mL; RRID: AB_772193
Donkey Anti-Rabbit IgG ECL Antibody, HRP Conjugated, GE Healthcare	GE Healthcare	Cat# NA9340-1mL; RRID: AB_772191
Mouse monoclonal anti-phospho (S139)-H2AX (JBW301)	Millipore Sigma	CAT# 05–636; RRID: AB_309864
Mouse anti-PAR/pADPr Monoclonal Antibody(Clone 10HA)	R & D Systems	Cat#4335-MC-100; RRID: AB_2572318
Histone H3 Rabbit Antibody #9715	Cell Signaling Technology	Cat# 9715S; RRID: AB_331563
Ub (P4D1) antibody, Santa Cruz Biotechnology	Santa Cruz Biotechnology	Cat# sc-8017; RRID: AB_628423
Anti-ds DNA antibody	Abcam	Cat# ab27156; RRID: AB_470907
TOP3A Polyclonal antibody	Proteintech	Cat#14525-1-AP; RRID: AB_2205881
FEN1 (E5X2T) Rabbit Monoclonal Antibody	Cell Signaling Technology	Cat#83104S
SPRTN Polyclonal Antibody	Invitrogen	Cat# PA5-110424; RRID: AB_2855835
PARP1 Antibody (F-2)	Santa Cruz Biotechnology	Cat# sc-8007; RRID: AB_628105
Goat anti-mouse Alexa Fluor 488	ThermoFisher Scientific	Cat# A-11001; RRID: AB_2534069
MRE11 antibody [12D7]	GeneTex	Cat# GTX70212; RRID: AB_372398
Bacterial and virus strains
NEB® 5-alpha Competent E. coli (High Efficiency)	New England BioLabs Inc.	Cat# C2987H
Chemicals, peptides, and recombinant proteins
DMEM - Dulbecco’s Modified Eagle Medium	ThermoFisher Scientific	Cat# 11965-092
Fetal Bovine Serum	Gemini	Cat# 100-106
Penicillin-Streptomycin	ThermoFisher Scientific	Cat# 15140-122
Trypsin-EDTA (0.05%)	ThermoFisher Scientific	Cat# 25300054
cOmplete Mini, EDTA-free (protease inhibitor cocktail)	Roche	Cat# 11836170001
Recombinant Human TOP3A	Gift from Gabor M. Harami, NHLBI, NIH	N/A
Lipofectamine 3000 Reagent	ThermoFisher Scientific	Cat# L3000015
Lipofectamine® RNAiMAX transfection reagent	ThermoFisher Scientific	Cat# 13778150
Benzonase	Sigma-Aldrich	Cat# E8263
Proteinase K	ThermoFisher Scientific	Cat# PI17916
Pierce™ ChIP-grade Protein A/G Magnetic Beads	ThermoFisher Scientific	Cat# 26162
Tris-glycine SDS sample buffer	Novex	Cat# LC2676
SuperSignal™ West Femto Maximum Sensitivity Substrate	ThermoFisher Scientific	Cat# 34095
DNAzol	ThermoFisher Scientific	Cat# 10503027
DMSO	Millipore Sigma	Cat# D8418
Talazoparib	MedChemExpress	Cat# HY-16106
PDD 00017273	Tocris Bioscience	Cat# 5952
TAK243	MedChemExpress	Cat# HY-100487
FEN1-IN-4	Selleck	Cat# S3397
Thymidine	Millipore Sigma	Cat# T1895
Vectasheild DAPI medium	VectorLabs	Cat# H-1200
N-Ethylmaleimide	Millipore Sigma	Cat# E3876-25G
Lenti-X Concentrator	TaKaRa, Japan	Cat# 631231
Paraformaldehyde (PFA) 20%	EMS	Cat# 15713-S
3M Sodium Acetate	Quality Biological	Cat# 351-035-721
Water, DEPC treated	Quality Biological	Cat# 351-068-721
RNase A	ThermoFisher Scientific	Cat# EN0531
Critical commercial assays
Duolink In Situ Detection Reagents Red	Millipore Sigma	Cat# DUO92008
Duolink In Situ PLA Probe Anti-Rabbit PLUS	Millipore Sigma	Cat# DUO92002
Duolink In Situ PLA Probe Anti-Mouse MINUS	Millipore Sigma	Cat# DUO92004
QuickChange II XL site-directed mutagenesis kit	Agilent Technologies	Cat# 200521
Click-iT EdU Alexa Fluor 488 Imaging Kit	Invitrogen	Cat# C10337
Deposited data
Original imaging data	This study: Mendeley Data	https://doi.org/10.17632/vgd8fb98vr.1
Experimental models: Cell lines
U2OS	Developmental Therapeutics Program, NCI/NIH	N/A
HCT116	Developmental Therapeutics Program, NCI/NIH	N/A
HEK193	ATCC	CRL-1573
MCF-7	Radiation Genetics Dept., Kyoto University, Japan	N/A
MCF-7 shFEN1	This study	N/A
LentiX293T	Clonetech, Japan	Cat# 632180
Oligonucleotides
SMARTpool: ON-TARGETplus FEN1 siRNA	Dharmacon	Cat# L-010344-00-0005
SMARTpool: ON-TARGETplus SPRTN siRNA	Dharmacon	Cat# L-015442-02-0005
SMARTpool: ON-TARGETplus PARP1 siRNA	Dharmacon	Cat# L-006656-03-0005
SMARTpool: ON-TARGETplus MRE11 siRNA	Dharmacon	Cat# L-009271-00-0005
ORF: ON-TARGETplus TOP3A siRNA	Dharmacon	Cat# L-005279-00-0005
Controlpool: ON-TARGETplus Control siRNA	Dharmacon	Cat# D-001206-13-5
Forward Primer for site directed mutagenesis of FEN1: CGAGGACATGGCCTGCCTCAC (5′-3′)	IDT	N/A
Reverse Primer for site directed mutagenesis of FEN1: GTGAGGCAGGCCATGTCCTCG (5′-3′)	IDT	N/A
shFEN1 Forward Primer for cloning into pLKO.1 CCGGGATGCCTCTATGAGCATTTATCTCGAGA! TAAATGCTCATA GAGGCATCTTTTTG (5′-3′)	Eurofin	N/A
shFEN1 Reverse Primer for cloning into pLKO.1 AATTCAAAAAGATGCCTCTATGAGCATTTATCT! CGAGATAAAT GCTCATAGAGGCATC (5′-3′)	Eurofin	N/A
Recombinant DNA
Human TOP3A-Myc-FLAG	OriGene	Cat# RC208236
Human FEN1-Myc-FLAG	OriGene	Cat# RC201785
pLKO.1	Addgene	Cat# 8453
pSPAX2	Addgene	Cat#12260
pMD2.G	Addgene	Cat#12259
Software and algorithms
GraphPad Prism 10 (software for drawing graphs and statistics analysis)	GraphPad Software	RRID: SCR_002798
Fiji	ImageJ	https://imagej.net/ij/
Image Lab 6.1	BIO-RAD	N/A
OpenComet	ImageJ Plugin (Fiji)	https://imagej.net/ij/

Highlights.

• The endonuclease FEN1 repairs DNA-protein crosslinks formed by topoisomerase IIIa (TOP3A-DPCs)

• Transient PARylation signals FEN1-mediated TOP3A-DPC repair

• PARP1 negatively regulates the ubiquitylation and SPRTN-/TDP2-mediated excision of TOP3A-DPCs